Classification of substances by health hazard using deep neural networks and molecular electron densities.

In this paper we present a method that allows leveraging 3D electron density information to train a deep neural network pipeline to segment regions of high, medium and low electronegativity and classify substances as health hazardous or non-hazardous. We show that this can be used for use-cases such as cosmetics and food products. For this purpose, we first generate 3D electron density cubes using semiempirical molecular calculations for a custom European Chemicals Agency (ECHA) subset consisting of substances labelled as hazardous and non-hazardous for cosmetic usage. Together with their 3-class electronegativity maps we train a modified 3D-UNet with electron density cubes to segment reactive sites in molecules and classify substances with an accuracy of 78.1%. We perform the same process on a custom food dataset (CompFood) consisting of hazardous and non-hazardous substances compiled from European Food Safety Authority (EFSA) OpenFoodTox, Food and Drug Administration (FDA) Generally Recognized as Safe (GRAS) and FooDB datasets to achieve a classification accuracy of 64.1%. Our results show that 3D electron densities and particularly masked electron densities, calculated by taking a product of original electron densities and regions of high and low electronegativity can be used to classify molecules for different use-cases and thus serve not only to guide safe-by-design product development but also aid in regulatory decisions. SCIENTIFIC CONTRIBUTION: We aim to contribute to the diverse 3D molecular representations used for training machine learning algorithms by showing that a deep learning network can be trained on 3D electron density representation of molecules. This approach has previously not been used to train machine learning models and it allows utilization of the true spatial domain of the molecule for prediction of properties such as their suitability for usage in cosmetics and food products and in future, to other molecular properties. The data and code used for training is accessible at https://github.com/s-singh-ivv/eDen-Substances .

Structure-Odor Relationships of ( Z)-3-Alken-1-ols, ( Z)-3-Alkenals, and ( Z)-3-Alkenoic Acids.

( Z)-3-Unsaturated volatile acids, alcohols, and aldehydes are commonly found in foods and other natural sources, playing a vital role in the attractiveness of foods but also as compounds with chemocommunicative function in entomology. However, a systematic investigation of their smell properties, especially regarding humans, has not been carried out until today. To close this gap, the odor thresholds in air and odor qualities of homologous series of ( Z)-3-alken-1-ols, ( Z)-3-alkenals, and ( Z)-3-alkenoic acids were determined by gas chromatography-olfactometry. It was found that the odor qualities in the series of the ( Z)-3-alken-1-ols and ( Z)-3-alkenals changed, with increasing chain length, from grassy, green to an overall fatty and citrus-like, soapy character. On the other hand, the odor qualities of the ( Z)-3-alkenoic acids changed successively from cheesy, sweaty via plastic-like, to waxy in their homologous series. With regard to their odor potencies, the lowest thresholds in air were found for ( Z)-3-hexenal, ( Z)-3-octenoic acid, and ( Z)-3-octenal.

Detection of Volatile Metabolites of Garlic in Human Breast Milk.

The odor of human breast milk after ingestion of raw garlic at food-relevant concentrations by breastfeeding mothers was investigated for the first time chemo-analytically using gas chromatography-mass spectrometry/olfactometry (GC-MS/O), as well as sensorially using a trained human sensory panel. Sensory evaluation revealed a clear garlic/cabbage-like odor that appeared in breast milk about 2.5 h after consumption of garlic. GC-MS/O analyses confirmed the occurrence of garlic-derived metabolites in breast milk, namely allyl methyl sulfide (AMS), allyl methyl sulfoxide (AMSO) and allyl methyl sulfone (AMSO2). Of these, only AMS had a garlic-like odor whereas the other two metabolites were odorless. This demonstrates that the odor change in human milk is not related to a direct transfer of garlic odorants, as is currently believed, but rather derives from a single metabolite. The formation of these metabolites is not fully understood, but AMSO and AMSO2 are most likely formed by the oxidation of AMS in the human body. The excretion rates of these metabolites into breast milk were strongly time-dependent with large inter-individual differences.